Will These 3 "Accelerated AI" Stocks Power The Next AI Boom?

In 2002, Akamai Technologies was trading for less than a single dollar. Over the years I followed, it would go to gain 16,497% and according to former hedge fund manager Jason Bodner, a major force behind the run was the internet's shift from copper connections to fiber optics. The old infrastructure couldn't keep up, but moving data with light helped break through that bottleneck and unlock an entirely new wave of growth. According to Bodner, this was the first acceleration. Then came the second acceleration with smartphones and Bodner says the same pattern repeated. Mobile networks struggled to handle the explosion in data. Once again, upgrading connections from copper to fiber helped open the door for what came next. He points to Cirrus Logic gaining more than 6,500% Lattice Semiconductor more than 12,000% Monolithic Power Systems more than 17,000% and Meta stock climbing 42-fold. Now, Bodner says we're entering the third acceleration. This time, it's AI. He claims data centers are hitting their own limits. Another shift from copper to light could unleash the next generation of stock market winners. Except this time, he believes the wealth ahead could dwarf every boom that came before it. And he claims he found three little-known stocks best positioned to capitalize. But their names are inside a report that comes with a $179 subscription to his research service. But I have good news. I went through his hour-long presentation and was able to figure out each stock based on the clues in the presentation. In this video, I'm going to reveal each one for free. There is multiple photonics stocks to cover in this presentation, so we're just going to jump right into it. Let's start first stock. Bodner describes a company supplying the networking technology that allows clusters of AI chips to communicate. He says two of the world's biggest AI spenders are its customers and Nvidia has tapped it as a partner. Then he gives us some numbers, sales growth of around 35% expectation for AI sales more than double in 2026 and a company roughly a tenth the size of Broadcom. All the clues point to Arista Networks, ticker A N E T. The revenue clue is a particularly strong march. Arista reported 35.1 sales growth in the first quarter of 2026. It also announced an AI network collaboration with Nvidia in May 2024. And Microsoft and Meta have both been major customers matching its clues about supplying some of the biggest spenders in AI. To understand the opportunity here, think about what happens when thousands of expensive AI chips have to work together. They need to constantly exchange enormous amounts of data. If the network gets congested, those chips can spend time waiting for information, reducing the value customers get from their investment. Arista sells the switches and software that helps manage the traffic and keep the network running efficiently. Its collaboration with Nvidia specifically addresses problems such as congestion, configuration, and coordinating the network with the computing hardware. A big part of Arista's appeal is its software. Its EOS operating system and CloudVision management platform help customers configure equipment, monitor performance, automate changes, and troubleshoot problems across their networks. In my view, the software and its integration into customers' operations help make Arista a valuable long-term supplier. Its connections to bond or its photonics theme comes through the networking systems that use optical links. Arista's newer platforms support optical connections designed to reduce power consumption alongside higher network speeds. It also serves broader cloud and enterprise network markets giving it several opportunities behind this particular AI pitch. And the financial results show why investors are so excited. In the second quarter of 2026, revenue reached approximately $3 billion up almost 38% from a year earlier. Its operating margin was around 45%. That means roughly 45 cents of every revenue dollar remained as operating profit after operating expenses. It's also finished the quarter with approximately $13.3 billion in cash and marketable securities. That combination of growth, profitability, and financial strength is a big reason I think Arista is a great company. The part that makes me hesitate, valuation. At around $187, Arista was trading at roughly 59 times trailing earnings and 40 times forward earnings estimates. Its market value is approximately $236 billion. The price already reflects investors thinking the stock is going to grow quite well into the future. Arista can keep expanding and still deliver a disappointing stock return if its results fall short of what investors have priced in or if investors become less willing to pay such a high earnings multiple. There's also customer concentration to consider. Its two largest customers accounted for a combined 42% of its revenue in 2025 and change in spending plans at either one can meaningfully affect results. So, my verdict is a borderline buy. I like the business. Arista has an established customer base, impressive profitability, and a credible opportunity as AI creates demand for faster, more sophisticated networks. But, I think the valuation is stretched at the moment and that limits how enthusiastic I can be about buying at today's price. Strong execution could justify the premium over time. A pullback or earnings growing into the valuation would make the opportunity more attractive for me. For now, the quality of the company earns its borderline buy rating while the price keeps it from being a stronger buy. >> I'm going to reveal the rest of the stocks in 15 seconds, but before I do, I want to tell you about my free report on the top 10 stocks to buy and hold right now. These are companies I believe have the best mix of strong long-term potential and growth. When you're done watching, click the link in the description, enter your email, and I'll send it straight to your inbox. Now, let's move on to the second stock that Bonner says has direct exposure to photonics. He says its hardware moves data as light between data centers and its annual sales exceed $6 billion and its earnings could double within two years. One company that fits several of those clues is Coherent, ticker CHR, although it could be Ciena as well. For this video, I'm going to say it's Coherent, but again, it could be Ciena. Coherent makes the lasers, optical components, and transceivers that allow information to travel through fiber optic connections. A transceiver converts electrical data into optical signals and back again. These devices help connect the server switches and other equipment inside data centers as well as networks spanning much longer distances. That gives Coherent a direct connection to the shift Bonner is describing. As AI systems expand, customers need to move more information between increasingly powerful chips. Coherent supplies components that help make those faster optical connections possible. Its portfolio includes products for connections with AI clusters, between racks, and across separate data centers. And Nvidia's backing the opportunity with substantial commitments. In March 2026, the companies announced a strategic agreement that included a multi-billion dollar purchase commitment and a separate $2 billion investment from Nvidia to support Coherent's research, manufacturing capacity, and operations. That gives us concrete evidence that Coherent's technology matters to the AI infrastructure buildout. The financial results also show this business gaining momentum. For fiscal 2026, Coherent generated approximately 7.1 billion in revenue, up about 23%. Its latest reported quarter was stronger still. Revenue reached approximately 2.05 billion, an an increase of roughly 34%. Adjusted earnings per share increased 74%, reaching 1.74 and its data center communication segment grew approximately 59% in that quarter, accounting for nearly 4/5 of total revenue. That segment includes both data centers and communication customers. What makes Coherent particularly interesting to me is the valuation is much better after its recent pullback. Its price is around $236 and the stock was trading around 25 times a forward earnings estimates. That compares with roughly 40 times forward earnings for Arista. On that measure, Coherent offers a substantially less demanding starting valuation. The shares were also 46% below their 52-week high of $440. So, this has been a substantial correction giving investors a much lower entry price than they would have paid near the highs. Now, there's still risks, especially with cash flow. In fiscal 2026, Coherent generated approximately 80 million in operating cash flow while spending around 1.1 billion on property, plant, and equipment. Its manufacturing expansion is consuming substantial cash and I want to see those investments translate into stronger cash generation. The forward valuation also depends on earnings estimates being achieved. Slower customer spending, production problems, and pressure on margins could weaken that investment case. But, weighing the risks against the growth opportunity and today's valuation, I'm putting Coherent in the buy category. I like its direct exposure to optical networking, the growth already showing up in the results, and the commercial relationship with Nvidia. And after this pullback, I think the price offers a more attractive balance between the opportunity and the expectation investors are paying for it. For someone investing over several years, I view this as a good buying opportunity in the AI infrastructure buildout. Now, let's figure out the third stock. Bodnar points to acquisitions expanding its a photonics capabilities, a multi-billion dollar relationship with Nvidia, and annual revenue growth of 42%. This is Marvell Technology, ticker M R V L. The 42% growth figure matches its fiscal 2026 results. And long-term viewers might remember this one. I recommended Marvell earlier this year on this channel, and the stock has exploded since then. I still think it's an excellent company, but after that run, getting a good entry price becomes much more important. Marvell designs several types of chips that help power AI infrastructure. These include custom processors developed for major cloud customers, networking chips that move information between systems, and chips used in high-speed optical connections. That gives it several ways to benefit as customers build larger, more sophisticated data centers. Its optical technology is particularly relevant to Bodnar's thesis. Marvell's signal processing chip help optical modules transmit large amounts of information reliably over fiber. As networks move towards higher speeds, those chips help manage increasingly demanding connections within and between data centers. The company has also been investing in what comes next. In February, Marvell completed its acquisition of Celestial AI, adding technology designed to connect large AI systems using optical links. That creates another potential growth opportunity, although Marvell expects initial revenue contribution from Celestial to begin in the second half of fiscal 2028. Investors still need to see that technology turn into meaningful commercial sales. Then there's Nvidia. In March, the companies announced a strategic partnership involving custom chips, networking, and silicon photonics. Nvidia also invested $2 billion in a Marvell. The agreement connects Marvell's technology with Nvidia's infrastructure through its NVLink Fusion platform, giving customers more flexibility when developing specialized AI systems. And the existing business is already growing rapidly. In its latest reported quarter, Marvell generated approximately $2.74 billion in revenue, up 37%. Data center revenue grew 46% and represented about 79% of total sales. The company also generated more than $600 million in operating cash flow during the quarter. At its October investor day, management raised its fiscal 2028 revenue outlook to approximately $20 billion, compared with its previous forecast of 18 billion. That gives you an idea of the growth management believes is ahead. These are forecasts, and delivering them will require successful customer deployments and strong execution. For me, the main issue is how much of that success investors are already paying for. The stock is around $244, and Marvell is trading around 43 times forward earnings estimates. That's a substantial premium, leaving the stock sensitive to changes in growth expectations. Large customer programs can experience delays, acquisitions can bring integration costs and execution risk, and a company can keep growing while its stock struggles if the results fall short of the expectations built into the valuation. So my preferred entry price would be around $200 or below. That's approximately 18% below the current price. Assuming earnings estimates remain unchanged, it would bring the forward earning multiple down from roughly 43 to about 35. For a business with Marvell's capabilities and growing opportunities, that's a premium I'd feel more comfortable paying. It would give me a better cushion against delays or results coming in below expectations. And before you go, don't forget to grab my free report on the top 10 stocks to buy and hold right now. These are companies I believe offer the best combination of long-term growth potential and strong underlying businesses. Just click the link in the description, enter your email, and I'll send the full report straight to your inbox.